United State. OpenAI introduces a new chatbot “milestone”.

According to the manufacturer, a new version of the chatbot ChatGPT aims to make the program, which is based on artificial intelligence, safer, more accurate and more like people.

GPT-4 is a “milestone” and the technology exhibits “human-level performance on certain professional and academic tasks,” California-based company OpenAI wrote in a blog post Tuesday when the new version was introduced.

ChatGPT generates texts with the help of artificial intelligence. Users can specify individual commands or sentences, which the system then completes independently with the help of massive amounts of data from the Internet. The new version of GPT-4 is “less good than a human in many real-life scenarios, but just as good as a human in many professional and academic contexts,” reads an OpenAI press release.

Image information can now also be used

According to the information, an important novelty compared to GPT-3.5 is that texts in the future must be generated from both text and image data. For example, if a user submits a picture of the inside of a refrigerator, GPT-4 will not only correctly identify what’s inside, but also suggest what to cook with the ingredients. However, the US company added that the functionality is not yet widely available.

Software company Microsoft, which has invested billions of dollars in the startup, announced Tuesday that the Bing search engine already works with GPT-4.

SDA